using noble gases to investigate mountain-front...
TRANSCRIPT
Using noble gases to investigate mountain-front recharge
Andrew H. Manning*, D. Kip Solomon
Department of Geology and Geophysics, University of Utah, Salt Lake City, UT 84112, USA
Received 26 February 2001; accepted 17 January 2003
Abstract
Mountain-front recharge is a major component of recharge to inter-mountain basin-fill aquifers. The two components of
mountain-front recharge are (1) subsurface inflow from the mountain block (subsurface inflow), and (2) infiltration from
perennial and ephemeral streams near the mountain front (stream seepage). The magnitude of subsurface inflow is of central
importance in source protection planning for basin-fill aquifers and in some water rights disputes, yet existing estimates carry
large uncertainties. Stable isotope ratios can indicate the magnitude of mountain-front recharge relative to other components,
but are generally incapable of distinguishing subsurface inflow from stream seepage. Noble gases provide an effective tool for
determining the relative significance of subsurface inflow, specifically. Dissolved noble gas concentrations allow for the
determination of recharge temperature, which is correlated with recharge elevation. The nature of this correlation cannot be
assumed, however, and must be derived for the study area. The method is applied to the Salt Lake Valley Principal Aquifer in
northern Utah to demonstrate its utility. Samples from 16 springs and mine tunnels in the adjacent Wasatch Mountains indicate
that recharge temperature decreases with elevation at about the same rate as the mean annual air temperature, but is on average
about 2 8C cooler. Samples from 27 valley production wells yield recharge elevations ranging from the valley elevation (about
1500 m) to mid-mountain elevation (about 2500 m). Only six of the wells have recharge elevations less than 1800 m. Recharge
elevations consistently greater than 2000 m in the southeastern part of the basin indicate that subsurface inflow constitutes most
of the total recharge in this area.
q 2003 Published by Elsevier Science B.V.
Keywords: Noble gases; Recharge; Temperature; Mountains; Ground water; Tracers
1. Introduction
Inter-mountain basin-fill aquifers are a major
source of groundwater in the western US. Existing
studies suggest that mountain-front recharge typically
accounts for one third to nearly all of the recharge to
these aquifers (Anderson and Freethey, 1996; Prudic
and Herman, 1996; Gates, 1995; Mason, 1998).
Mountain-front recharge consists of two components:
(1) subsurface inflow from the mountain block
(subsurface inflow), and (2) infiltration from perennial
and ephemeral streams near the mountain front
(stream seepage) (Anderson et al., 1992). Neither
component consistently dominates. While stream
seepage can be estimated using gauging data or
seepage studies, estimates of the magnitude and
distribution of subsurface inflow remain very poorly
constrained. Estimates based on the disposal of
Journal of Hydrology 275 (2003) 194–207
www.elsevier.com/locate/jhydrol
0022-1694/03/$ - see front matter q 2003 Published by Elsevier Science B.V.
doi:10.1016/S0022-1694(03)00043-X
* Corresponding author. Present address; U.S. Geological Survey,
P.O. Box 25046, Mail Stop 973 Denver, CO 80225–0046, USA.
Tel.: +1–303–236–1812; fax: +1–303–236–3200.
E-mail address: [email protected] (A.H. Manning).
mountain precipitation (e.g. Chavez et al., 1994) are
hampered by large uncertainties in evapotranspiration
and snowpack sublimation rates. Estimates based on
calibrated steady-state groundwater flow models of
the basin-fill aquifers (e.g. Lambert, 1995), are often
highly non-unique due to a scarceness of hydrogeo-
logic data near the mountain front and poorly
constrained discharge rates. When subsurface inflow
estimates are large they seem at odds with the
potential for range-bounding normal faults to act as
barriers to subsurface inflow (Caine et al., 1996) and,
in some cases, the apparently low permeability of
rocks in the adjacent mountains (e.g. Parry et al.,
2000). The inability to confidently assess the relative
magnitude of subsurface inflow presents a significant
problem for those charged with source protection
planning for valley wells, and for those involved in
water rights disputes in rapidly developing mountain
resort areas.
Dissolved noble gas concentrations potentially
provide an excellent tool for identifying subsurface
inflow because of their dependence on the ground
temperature at the point of recharge. For typical water
table depths (5–100 m), the ground temperature near
the water table, and thus the noble gas recharge
temperature, is approximately equal to the mean
annual air temperature at the surface above (e.g.
Mazor, 1991; Stute and Sonntag, 1992). The decrease
in mean annual air temperature with elevation
(atmospheric lapse) should result in subsurface inflow
having a lower noble gas recharge temperature than
water recharged in the valley. Oxygen-18 and 2H data
have been effectively used to distinguish mountain-
front recharge from other sources of recharge (e.g.
Thiros, 1995). However, stable isotope ratios are
generally incapable of distinguishing subsurface
inflow from stream seepage because both sources
carry a depleted, high elevation signal. Therefore,
d 18O and d 2H values can be used to specifically
identify subsurface inflow only in those rare cases
when stream seepage is known with considerable
confidence to be negligible. The recharge temperature
of stream seepage, however, should be that of other
water recharged in the valley because water table
depths beneath losing streams near the margins of
inter-mountain basins are typically greater than 30 m.
In this study, the possibilities and limitations of using
noble gases as tracers of subsurface inflow are
explored theoretically, then the method is applied to
the Salt Lake Valley Aquifer in northern Utah to
confirm its utility.
2. Noble gases as recharge elevation tracers
The possibility of using noble gases as recharge
elevation tracers either by themselves or in conjunc-
tion with 18O and 2H has been discussed by previous
workers (Mazor, 1991; Rauber et al., 1991; Zuber
et al., 1995; Aeschbach-Hertig et al., 1999; Ballentine
and Hall, 1999). However, a rigorous application of
the method has not been reported. The general form of
the governing equation for the concentration of noble
gas i dissolved in fresh water ðCiÞ is:
Ci ¼ CEi þ CA
i ð1Þ
where CEi is the component resulting from equili-
bration with the atmosphere and CAi is the component
resulting from excess air. Excess air (dissolved
atmospheric gases in excess of equilibrium values)
is ubiquitous in groundwater (Heaton and Vogel,
1981) and is commonly attributed to dissolution of air
bubbles trapped during water table rises (Stute and
Schlosser, 2000). The equilibrium component is
related to the dry atmospheric pressure at the recharge
point (recharge pressure, Pr) and to the temperature at
the water table at the recharge point (recharge
temperature, Tr) by:
CEi ¼
XiPr
KiðTrÞð2Þ
where Xi is the dry mole fraction of gas i in air and Ki
is the Henry’s Law constant for gas i; a nonlinear
function of Tr of the form:
1
Ki
¼ exp ai
bi
Tr
2 1
� �þ 36:855
bi
Tr
2 1
� �2" #
ð3Þ
where ai and bi are empirical constants (Solomon
et al., 1998). The excess air component is given by:
CAi ¼ XiA ð4Þ
where A is the total concentration of excess air.
Aeschbach-Hertig et al. (2000) have recently pro-
posed more complex forms of Eq. (4) that probably
model the formation of excess air more accurately.
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207 195
However, these were not employed in this study
mainly because the same authors also demonstrated
that the use of different excess air models does not
alter the derived Tr by more than about 1 8C. The
relationship between Pr and the recharge elevation in
meters ðHÞ can be closely approximated with:
Pr ¼ exp 2H
8300
� �2 PH2O ð5Þ
where PH2O is the partial pressure of water, provided
H , 3000 m (Plummer and Busenberg, 2000).
If three or more dissolved gas concentrations are
measured, then a system of equations in the form of
(1) can be in effect solved simultaneously using
standard inverse techniques to determine the unknown
parameters Tr; H; and A: Concentrations of He, Ne,
Ar, Kr, and Xe are typically measured, though CHe is
seldom used because the addition of terrigenic 4He
after recharge is common. Aeschbach-Hertig et al.
(1999) and Ballentine and Hall (1999) both employ
computer codes that solve this inverse problem by
minimizing x2; defined as:
x2 ¼X
i
ðCi 2 Cmodi Þ2
s2i
ð6Þ
where Ci are the measured concentrations, Cmodi are
the modeled concentrations, and s2i are the exper-
imental 1s errors. These authors discuss this approach
in detail. A similar x2 minimization code was written
for this study to quantify uncertainty in the derived
parameters, given (a) the measurement uncertainties
particular to our sampling and analytical process, and
(b) the use of N2 instead of Xe (use of N2 as a
conservative gas in this study is warranted because
groundwater in the study area is generally well
oxygenated, precluding appreciable denitrification).
Our inversion code uses the Newton method (Zhda-
nov, 2000) to perform the minimization because the
problem is non-linear, over-determined, and has only
three parameters and four measurements (CNe; CAr;
CKr; CN2).
The dependence of Ci on H and Tr presents the
possibility of using noble gases as recharge elevation
tracers either directly through the derivation of H; or
indirectly by deriving Tr and using the
local atmospheric lapse rate. Unfortunately, when
Tr; H; and A are all unknown, the inverse problem is
ill-posed; given typical measurement errors of 1%
(1sÞ; uncertainties for derived values of H and Tr are
unacceptably large (Aeschbach-Hertig et al., 1999;
Ballentine and Hall, 1999). The results of Monte
Carlo simulations clearly demonstrate this behavior
(Fig. 1a). In these simulations, statistically plausible
samples were generated for a given set of recharge
conditions (specified Tr; H; and A). The randomly
assigned errors in Ci were normally distributed with
an assigned si determined from the analysis of sample
replicates in our laboratory (3% for Kr, 2% for the
remaining gases). Values of Tr; H; and A were then
derived for each synthetic sample and the s of the
derived distribution of each parameter was considered
the uncertainty for that parameter. Several reasonable
recharge conditions were simulated. Derived uncer-
tainties in H and Tr were typically around ^1500 m
and ^6 8C, respectively. Because the potential
recharge elevation range in most studies is
,2000 m, and typical lapse rates are 25 to 28 8C
per kilometer of elevation gain, an accuracy of about
^200 m in H or ^1 8C in Tr would be required to put
useful constraints on recharge elevation. To achieve
this, a 1s measurement error of #0.3% for all gases
would be required, which is beyond our present
analytical capabilities.
The inverse problem becomes well-posed, how-
ever, when H is known a priori, as is the case when the
method is applied to investigate paleoclimate (Aesch-
bach-Hertig et al., 1999; Ballentine and Hall, 1999).
Monte Carlo simulations like those described above
were performed for this situation and the uncertainty
in Tr decreased to ^0.9 8C (Fig. 1b), given the
measurement uncertainties for our laboratory. This
level of accuracy was confirmed by tests performed
using water equilibrated with air under controlled
conditions such that Tr; H; and A were known (Fig. 2).
The x2 surface was mapped for several synthetic
samples and all displayed a correctly located, well-
defined minimum point with no local minima (Fig. 3).
The solution to this inverse problem thus appears to be
both unique and stable.
Zuber et al. (1995) and Aeschbach-Hertig et al.
(1999) point out that in mountainous terrain where H
is initially unknown, the noble gas data can be used to
derive a set of best-fit H –Tr pairs for a given sample
by specifying various values of H (Fig. 4). Assuming
Tr is equal to the mean annual air temperature ðTaÞ at
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207196
Fig. 1. Typical results of Monte Carlo simulations performed to determine uncertainty in derived recharge parameters. The two simulations
shown each involved 1000 realizations of an alpine water with recharge temperature ðTrÞ ¼ 7 8C, recharge elevation ðHÞ ¼ 2000 m, and excess
air ðAÞ ¼ 0.002 cm3 STP/g. (a) Case where Tr; H; and A are all unknown. The span of the cluster of solutions over significant ranges of Tr; H;
and A indicates large uncertainties for all three parameters. Uncertainties (1sÞ are ^6.3 8C for Tr; ^1650 m for H; and ^0.00124 cm3 STP/g
for A: (b) Case where H is known a priori. The modest range of solutions indicates acceptable uncertainties for Tr and A (^0.8 8C and
^0.00025 cm3 STP/g, respectively, for this particular water).
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207 197
all elevations (implying that the water table is near
land surface), the actual H and Tr for the sample
should be the point of intersection between the linear
set of best-fit solutions and a line representing the
local atmospheric lapse on a plot of H versus Tr:
These authors apply this approach to a small number
of samples and obtain mixed results; some derived H
values are reasonable while others are difficult to
explain.
This study in effect constitutes a considerably more
rigorous application of the approach described above.
A significant modification, however, is that Tr is not
assumed to equal Ta for mountain groundwater flow
systems. At high elevation, snowmelt is often the main
source of recharge. Large volumes of snowmelt water
infiltrating rapidly through fractured rock could
conceivably remain near 0 8C in the unsaturated
zone, particularly since water table depths may
decrease to ,3 m during periods of snowmelt (Hill,
1990; Buttle, 1989). Hence, Tr could be less than Ta in
many alpine areas. However, mean annual surface
ground temperatures extrapolated from borehole
temperature profiles are commonly about 3 degrees
warmer than Ta (Powell et al., 1988), and ground
temperatures several degrees warmer than Ta have
been observed in areas of prolonged snow cover
(Smith et al., 1964). Furthermore, mountain water
tables can be as deep as 500 m in more arid regions of
the western US (Fridrich et al., 1994). Water table
temperatures in these areas may be well above Ta if
recharge rates are not large enough to significantly
depress the geothermal gradient. Given these possi-
bilities, the local relationship between H and Tr was
derived in this study by determining Tr for mountain
groundwaters with constrained values of H:
3. Example application: mountain-front recharge
to the eastern Salt Lake Valley
3.1. Site description
The Salt Lake Valley is an active rift basin located
in northern Utah (Fig. 5). It is semiarid with a mean
annual precipitation of 30 to 50 cm, depending on
location, and a mean annual air temperature of about
12 8C. The basin is at an elevation of 1300 to1500 m,
and is bound on the east and west by mountains that
rise to over 2700 m. This study focuses on mountain-
front recharge to the eastern side of the basin from the
central Wasatch Mountains (Fig. 5). Mean annual
precipitation in the Wastach Mountains is 50 to
130 cm, depending mainly on elevation, and most
precipitation falls as snow.
The basin-fill material consists of mostly-consoli-
dated Tertiary sediments overlain by mostly-uncon-
solidated Quaternary sediments (Hely et al., 1971).
Nearly all wells in the valley are completed in the
Quaternary sediments that are 120–300 m thick
throughout most of the basin. Production wells are
screened within the Principal Aquifer, the deeper
Quaternary sediments composed of sand and gravel
inter-bedded with lenses of fine-grained deposits. The
fine-grained lenses increase in frequency and thick-
ness with distance from the mountains, meaning that
Fig. 2. Results of accuracy tests performed on water equilibrated
with air at (a) 25 8C and (b) 7 8C. Each point is a separate sample
collected from the same batch of air-equilibrated water. Solid lines
show the known Tr; which has an uncertainty of ^0.5 8C (shown by
dashed lines) due to variations in room temperature during
equilibration. Error bars indicate uncertainty of ^0.9 8C in derived
Tr based on Monte Carlo simulations.
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207198
the Principal Aquifer is generally unconfined near the
mountains and confined elsewhere in the basin.
3.2. Existing knowledge of mountain-front recharge
Groundwater flow models of the basin indicate that
subsurface inflow is significant, constituting 45% of
the total recharge, while stream seepage only accounts
for 5% of the total recharge (Waddell et al., 1987;
Lambert, 1995). Calculations based on the disposal of
mountain precipitation corroborate this large subsur-
face inflow estimate (Hely et al., 1971). However, a
careful error analysis performed for this study
indicates that the estimate has an uncertainty of
^50 to 100%. Further, the apparently low per-
meability of rocks in large areas of the central
Wasatch range seem to preclude such large subsurface
inflow rates (Parry et al., 2000). Uncertainty in the
stream seepage estimate is also significant because it
is based on stream gauge data that is complicated by
numerous unmeasured inputs and diversions (Hely
et al., 1971). The remainder of the recharge budget
Fig. 3. Misfit surface ðlog10 x2Þ for an alpine water with recharge temperature ðTrÞ ¼ 7 8C, recharge elevation ðHÞ ¼ 2000 m, and excess air
ðAÞ ¼ 0.002 cm3 STP/g. The misfit surface was mapped at increments of 0.2 8C and 0.0001 cm3 STP/g excess air. The stability and uniqueness
of the derived solution is demonstrated by the fact that the surface is smooth with a single, well-defined minimum point properly located at the
actual parameter values.
Fig. 4. Method of determining recharge elevation ðHÞ from noble
gas concentrations proposed by Zuber et al. (1995) and Aesch-
bach-Hertig et al. (1999). A family of H –Tr solutions can be
derived from the noble gas data by specifying different values of H:
A unique recharge elevation is obtained at the intersection between
the curve representing this family of H –Tr solutions and a line
representing the local atmospheric lapse (mean annual air
temperature vs. elevation).
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207 199
mainly consists of seepage from precipitation (20%),
unused irrigation water (20%), and canal water (10%)
in the valley (Hely et al., 1971; Waddel et al., 1987;
Lambert, 1995). This seepage in combination with
stream seepage will henceforth be referred to as
‘valley recharge’. Valley recharge is believed to occur
largely near the mountain front where the principal
aquifer is unconfined.
3.3. Constraining recharge temperatures of source
waters
Noble gas samples were collected from 16 springs
and mine tunnels located at various elevations on the
western slope of the Wasatch Mountains. Samples
were collected using passive diffusion samplers
similar to those shown in Sanford et al. (1996).
Samplers were placed directly within the spring
orifice to insure that the sampled water had not re-
equilibrated with the atmosphere. The samplers were
allowed to equilibrate for at least 24 h. Nitrogen, Ne,
Ar, and Kr were measured using a quadrapole mass
spectrometer. Samples were first inlet directly into a
high vacuum purification system that includes a Ti/Zr
sponge for reactive gases. Non-reactive gases were
cryogenically separated. Nitrogen and O2 were
measured dynamically using a leak valve, while the
remaining gases were measured statically.
Fig. 5. Location of study area. The Jordan River flows northward toward the Great Salt Lake and is the primary discharge zone for the Salt Lake
Valley Principal Aquifer.
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207200
From the measured concentrations of Ne, Ar, Kr,
and N2, a maximum and a minimum Tr (Trmax and
Trmin) were derived for each sample using minimum
and maximum H values (Hmin and Hmax), respectively
(Fig. 6). The discharge elevation was used for Hmin
and the highest local surface water drainage divide
was used for Hmax: If Hmin or Hmax yielded an
unacceptably high x2 value (,5% probability of
producing measured concentrations), the parameter
was adjusted until an acceptable fit was achieved,
further constraining the recharge elevation. The local
atmospheric lapse of 26.7 8C/km shown in Fig. 6 is
that reported by Hely et al. (1971) for the Salt Lake
City area. This is close to the regional lapse for central
Utah (26.9 8C/km) reported by Powell et al. (1988).
Discharge temperatures lie near the atmospheric
lapse, most being slightly warmer (0 to 1 8C) than
Ta: Nearly all derived values of Tr; however, are 0 to
4 8C less than Ta: The least squares regression line
through the derived Tr values (including all Trmax and
Trmin values) has a slope of 27.3 8C/km, similar to the
atmospheric lapse, and indicates that Tr is on average
about 2 8C less than Ta: Clearly, assuming that Tr ¼ Ta
in the central Wasatch Mountains is inappropriate.
Fig. 6 also shows the surface ground temperature at an
elevation of approximately1850 m deduced from
.50 bore hole temperature profiles measured at the
Jordanelle Dam site (Moran, 1991). Both the mean
and the scatter of the surface ground temperatures is in
good agreement with derived Tr values.
The possible H –Tr combinations for springs and
tunnels shown in Fig. 6 define a zone of possible
H –Tr combinations for subsurface inflow (Fig. 7).
The zone of possible H –Tr combinations for valley
Fig. 6. Measured discharge temperature and derived recharge temperature ðTrÞ vs. elevation for springs and mine tunnel discharge
waters in the central Wastach Mountains. Discharge temperature is plotted against discharge elevation. Each pair of Tr values
connected with a tie line corresponds to a single sample; upper point ¼ (Trmin; Hmax) and lower point ¼ (Trmax; Hmin), where H is
recharge elevation. The linear regression line for all derived Tr values (including all Trmax and Trmin values) and the local atmospheric
lapse from Hely et al. (1971) are also shown. Jordanelle Dam site data from Moran (1991) consists of .50 mean annual surface
ground temperatures determined from borehole temperature profiles. The mean value is shown, with error bars indicating the
approximate range of scatter.
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207 201
recharge (Fig. 7) was determined from by tempera-
ture measurements near the water table in valley
wells. Measurements from this study are from wells
with water table depths of 10 to 30 m located in
recharge areas. Measurements from Thiros (1995)
are from shallow wells (,30 m) with unspecified
water table depths located throughout the eastern
side of the valley. The measured water table
temperatures generally range from 12 to 15 8C, as
expected given that shallow water table tempera-
tures typically exceed the local Ta by 1 to 2 8C
(Domenico and Schwartz, 1990). Furthermore, a Tr
of 13.9 8C was obtained from a water table well in
the valley where an 18O measurement verified local
recharge of evaporated irrigation canal water. A line
representing derived H –Tr combinations for a
hypothetical sample containing 100% valley
recharge (‘sample H –Tr line’) is also shown on
Fig. 7. The fact that this line intersects the zone of
possible H –Tr combinations for valley recharge, but
misses the zone of possible H –Tr combinations for
subsurface inflow, indicates that this sample could
be identified as valley recharge water by its noble
gas concentrations. Because only a small fraction of
sample H –Tr lines would intersect both zones of
possible H –Tr combinations, the method appears
capable of distinguishing subsurface inflow from
valley recharge.
However, given the long screens in the production
wells sampled for this study, samples containing
mixtures of subsurface inflow and valley recharge are
possible, if not probable. Two-component mixing
between waters with various recharge parameters was
simulated using synthetic samples. In all cases, the Tr
derived for the mixture was within 0.5 8C of the
weighted average of the end-member Tr values.
Therefore, mixing is approximately linear; i.e.
mixed samples lie near a line connecting the
Fig. 7. Zones of possible recharge elevation-temperature ðH –TrÞ combinations for subsurface inflow and valley recharge. The point
representing data from Thiros (1995) is the mean for 19 water table temperature measurements, the error bars indicating the approximate
data scatter. The line corresponding to a hypothetical sample containing 100% valley recharge does not intersect the zone of possible
H –Tr combinations for subsurface inflow. This demonstrates that the method is capable of distinguishing valley recharge from
subsurface inflow.
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207202
end-member waters in H –Tr space. This means that
the corners of the two zones of possible H –Tr
combinations shown in Fig. 7 can be connected to
form a single zone of possible H –Tr combinations for
the entire system (Fig. 8). This ‘solution zone’
contains all possible end-member waters and mixtures
thereof. Similarly, the ends of the dashed best-fit lines
in Fig. 7 can be connected to form a zone of probable
H –Tr combinations for the entire system. This
‘probable solution zone’ contains the most probable
end-member waters and their mixtures. When a
sample H –Tr line is plotted in H –Tr space, the
sample’s Hmin; Hmax; Trmin; and Trmax are all defined
by its intersection with the solution zone (Fig. 8). Note
that if the well elevation is greater than the elevation
of the point where the sample H –Tr line intersects the
bottom of the solution zone, the former is chosen as
the sample’s Hmin (and Trmax is chosen accordingly).
Also note that if Trmax is less than 11 8C, the sampled
water must contain at least some subsurface inflow.
The probable recharge elevation ðHprobÞ and the
probable recharge temperature ðTrprobÞ are defined by
the point of intersection between the sample H –Tr
line and the line bisecting the probable solution zone
(Fig. 8). Finally, the minimum fraction of subsurface
inflow in a sample is defined by the point where the
sample H –Tr line intersects the mixing line between a
typical valley recharge water and subsurface inflow
water with Tr ¼ 0 8C:
3.4. Valley well results
Noble gas samples were collected from 27
valley production wells screened in the Principal
Aquifer (Fig. 9). Total well depths are typically
100–200 m, and screen lengths are generally 50–
100 m. Flow rates of 500–2000 gallons per minute
(32–126 l/s) are typical. Well S25 (southernmost
Fig. 8. Zones of possible and probable recharge elevation–temperature ðH –TrÞ combinations for the entire system (includes mixed waters).
Hmin; Trmax; etc. are defined by points of intersection between the sample H –Tr line and different parts of these zones (see text for detailed
explanation).
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207 203
well on Fig. 9) is the only exception; it is a smaller
artesian well screened in the mountain block and is
not pumped. Samples were collected while the
wells were operating, and most wells had been
pumping for at least 48 h prior to sampling. As
with the springs, passive diffusion samplers were
used. The samplers were placed within a flow-
through cell plumbed directly into the well head.
The well pump provided significant pressure in the
flow-though cell, preventing gas loss. Flow rates
through the cell were maintained at about 4 l/min
throughout the 24-h (minimum) equilibration
period. Analytical procedures were the same as
described above.
For each sample, Trmax and Hprob were derived
using measured concentrations of Ne, Ar, Kr, and N2
(Figs. 10 and 11). Derived Trmax values range from 5.9
to 11.4 8C. All but two Trmax values are less than
11 8C, the minimum possible recharge temperature for
valley recharge. Therefore, water throughout most of
the study area is composed of at least some subsurface
inflow. A prominent cell of low-Trwater is located in
Fig. 9. Location of sampled wells. Potentiometric surface contours
for the Principal Aquifer are also shown to illustrate the general
direction of groundwater flow.
Fig. 10. Distribution of derived maximum possible recharge
temperature ðTrmaxÞ in the Principal Aquifer. The probable recharge
temperature is 3 to 4 8C colder for wells with Trmax , 7 8C:
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207204
the southern part of the study area. The Trmax values of
6 to 9 8C within this cell correspond to minimum
subsurface inflow fractions of 0.85 to 0.47, respect-
ively. Note that well S25, known to contain 100%
subsurface inflow, has a Trmax of 5.9 8C, similar to
Trmax values in the central part of the cell. Derived
Hprob values range from 1460 to 2580 m. Within the
low-Tr cell, Hprob values are 2000 to 2500 m, well
above the valley elevation. The fact that the core of
the low-Tr cell is located out away from the range
front suggests that much of the subsurface inflow does
not simply enter the Principal Aquifer laterally at the
range front, as generally envisioned by previous
workers. Instead, a large fraction apparently enters the
aquifer well out into the valley. Derived Trmax values
approach 11 8C in some areas outside of the low-Tr
cell. Water in these areas may be composed
dominantly of valley recharge.
4. Conclusions
Noble gas concentrations provide a useful tool
for investigating mountain-front recharge. While
d 18O data generally cannot be used to distinguish
between subsurface inflow and stream seepage,
recharge temperatures derived from noble gas data
put meaningful constraints on the relative signifi-
cance of subsurface inflow, specifically. The
general method of determining recharge elevation
from noble gas concentrations proposed by Zuber
et al. (1995) and Aeschbach-Hertig et al. (1999) is
feasible, with one significant modification: the local
relationship between Tr and H must be initially
determined because the assumption that Tr ¼ Ta in
mountain groundwater flow systems is dubious.
Application of the method to the Salt Lake Valley
Principal Aquifer yielded reasonable results. The
noble gas data clearly indicate that subsurface
inflow is significant in the eastern part of the basin,
and reveal its general distribution. A prominent cell
of low-Tr water is present in the southeastern part
of the valley. Water in this area is
composed dominantly, if not entirely, of subsurface
inflow. The results of this study lend considerable
weight to the large subsurface inflow rates
estimated for the Salt Lake Valley in previous
studies.
Acknowledgements
This study was funded in part by Salt Lake City
Public Utilities. We thank Sue Thiros of the U.S.G.S.
for her help with sampling logistics and general
feedback. We also thank the following agencies for
providing access to their wells, and for all of their
Fig. 11. Distribution of derived probable recharge elevation ðHprobÞ
in the Principal Aquifer. Values greater than 2000 m represent
greater than 50% subsurface inflow.
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207 205
assistance during our sampling effort: Salt Lake City
Public Utilities, Murray City Public Services, Jordan
Valley Water Conservancy District, Sandy City
Public Utilities, City of South Salt Lake Public
Works, City of Midvale Public Works, White City
Water Improvement District, and the Holladay Water
Company.
References
Aeschbach-Hertig, W., Peeters, F., Beyerle, U., Kipfer, R., 1999.
Interpretation of dissolved atmospheric noble gases in natural
waters. Water Resour. Res. 35, 2779–2792.
Aeschbach-Hertig, W., Peeters, F., Beyerle, U., Kipfer, R., 2000.
Paleotemperature reconstruction from noble gases in ground
water taking into account equilibration with entrapped air.
Nature 405, 1040–1043.
Anderson, T.W., Freethey, G.W., 1996. Simulation of ground-
water flow in alluvial basins in south-central Arizona and
parts of adjacent states. US Geol. Surv. Professional Pap.,
1406-D.
Anderson, T.W., Freethey, G.W., Tucci, P., 1992. Geohydrology
and water resources of alluvial basins in south-central Arizona
and parts of adjacent states. U.S. Geol. Surv. Professional Pap.,
1406-B.
Ballentine, C.J., Hall, C.M., 1999. Determining paleotemperature
and other variables by using an error-weighted, nonlinear
inversion of noble gas concentrations in water. Geochim.
Cosmochim. Acta 63, 2315–2336.
Buttle, J.M., 1989. Soil moisture and groundwater responses to
snowmelt on a drumlin sideslope. J. Hydrol. 105, 335–355.
Caine, J.S., Evans, J.P., Forster, C.B., 1996. Fault zone architecture
and permeability structure. Geology 24, 1025–1028.
Chavez, A., Davis, S.N., Sorooshian, S., 1994. Estimation of
mountain front recharge to regional aquifers: 1. Development of
an analytical hydroclimatic model. Water Resour. Res. 30,
2157–2167.
Domenico, P.A., Schwartz, F.W., 1990. Physical and Chemical
Hydrogeology, Wiley, New York.
Fridrich, C.J., Dudley, J.r., W, W., Stuckless, J.S., 1994.
Hydrogeologic analysis of the saturated-zone ground-water
system, under Yucca Mountain, Nevada. J. Hydrol. 154,
133–168.
Gates, J.S., 1995. Description and quantification of the ground-
water basins of the Wasatch Front, Utah, 1904–1994. In: Lund,
W.R., (Ed.), Environmental and Engineering Geology of the
Wasatch Front Region, Utah Geol. Surv. Publication 24, Salt
Lake City, UT, pp. 221–248.
Heaton, T.H.E., Vogel, J.C., 1981. ‘Excess air’ in groundwater.
J. Hydrol. 50, 201–216.
Hely, A.G., Mower, R.W., Harr, C.A., 1971. Water resources of Salt
Lake County, Utah. Utah Department of Natural Resources
Technical Publication No. 31.
Hill, B.R., 1990. Groundwater discharge to a headwater valley,
northwestern Nevada, USA. J. Hydrol. 113, 265–283.
Lambert, P.M., 1995. Numerical simulation of ground-water flow
in basin-fill material in Salt Lake Valley, Utah. Utah
Department of Natural Resources Technical Publication No.
110-B.
Mason, J.L., 1998. Ground-water hydrology and simulated effects
of development in the Milford area, an arid basin in south-
western Utah. US Geol. Surv. Professional Pap., 1409-G.
Mazor, E., 1991. Applied Chemical and Isotopic Groundwater
Hydrology, Halsted Press, New York.
Moran, K.J., 1991. Shallow thermal regime at Jordanelle Dam site,
central Rocky Mountains, Utah. Master’s Thesis, University of
Utah, Salt Lake City, UT.
Parry, W.T., Forster, C.B., Solomon, D.K., James, L.P., 2000.
Ownership of mine-tunnel discharge. Ground Water 38,
487–496.
Plummer, L.N., Busenberg, E., 2000. Chlorofluorocarbons. In:
Cook, P.G., Herczeg, A.L. (Eds.), Environmental Tracers in
Subsurface Hydrology, Kluwer Academic Publishers, Boston,
pp. 441–478.
Powell, W.G., Chapman, D.S., Balling, N., Beck, A.E., 1988.
Continental heat-flow density. In: Haenel, R., Rybach, L.,
Stegena, L. (Eds.), Handbook of Terrestrial Heat-Fow Density
Determination, Kluwer Academic Publishers, Boston, pp.
167–222.
Prudic, D.E., Herman, M.E., 1996. Ground-water flow and
simulated effects of development in Paradise Valley, a basin
tributary to the Humboldt River in Humboldt County, Nevada.
US Geol. Surv. Professional Pap., 1409-F.
Rauber, D., Loosli, H.H., Schmassmann, H., Andrews, J.N., 1991.
Noble gases in groundwater. In: Pearson, F.J. Jr., Balderer, W.,
Loosli, H.H., Lehmann, B.E., Matter, A., Peters, Tj., Schmass-
mann, H., Gautschi, A. (Eds.), Applied Isotope Hydrology: A
Case Study in Northern Switzerland, Elsevier, Amsterdam, pp.
116–152.
Sanford, W.E., Shropshire, R.G., Solomon, D.K., 1996. Dissolved
gas tracers in groundwater: simplified injection, sampling, and
analysis. Water Resour. Res. 32, 1635–1642.
Smith, G.D., Newhall, F., Robinson, L.H., Swanson, D., 1964. Soil
temperature regimes: their characteristics and predictability. US
Soil Conserv. Serv. Rept. SCS-TP-144.
Solomon, D.K., Cook, P.G., Sanford, W.E., 1998. Dissolved gases
in subsurface hydrology. In: Kendall, C., McDonnell, J.J. (Eds.),
Isotope Tracers in Catchment Hydrology, Elsevier, Amsterdam,
pp. 291–318.
Stute, M., Schlosser, P., 2000. Atmospheric noble gases. In: Cook,
P.G., Herczeg, A.L. (Eds.), Environmental Tracers in Subsur-
face Hydrology, Kluwer Academic Publishers, Boston,
pp. 349–377.
Stute, M., Sonntag, C., 1992. Paleotemperatures derived from noble
gases dissolved in groundwater in relation to soil temperature.
In: IAEA Staff, (Ed.), Isotopes of Noble Gases as Tracers in
Environmental Studies, IAEA, Vienna, pp. 111–122.
Thiros, S.A., 1995. Chemical composition of ground water,
hydrologic properties of basin-fill material, and ground water
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207206
movement in Salt Lake Valley, UT. Utah Department of Natural
Resources Technical Publication No. 110-A.
Waddell, K.M., Seiler, R.L., Santini M., Solomon, D.K., 1987.
Ground-water conditions in Salt Lake Valley, Utah, 1969–
1983, and predicted effects of increased withdrawals from wells.
Utah Department of Natural Resources Technical Publication
No. 87.
Zhdanov, M.S., 2000. Geophysical Inversion Theory and
Applications, Lecture Notes, University of Utah, Salt Lake
City, UT.
Zuber, A., Weise, S.M., Osenbruck, K., Grabczak, J., Ciezkowski,
W., 1995. Age and recharge area of thermal waters in Ladeck
Spa (Sudeten, Poland) deduced from environmental isotope and
noble gas data. J. Hydrol. 167, 327–349.
A.H. Manning, D.K. Solomon / Journal of Hydrology 275 (2003) 194–207 207